While air utilization efficiency was substantially improved by intermittent aeration in hollow fiber membranes in addition to the module design optimization, it has been improved mainly by stacking up membrane panels and optimizing bubble size in flat sheet membranes. Although there is a limit in stacking up due to the bubble coalescence and the subsequent deterioration of scouring effect, cassette height has been raised over time. The optimum bubble size is seemingly dependant on condition, but it is in decreasing trend for commercial membranes.
Submerged flat sheet MBRs use nozzles of different sizes, for example, the Kubota MBR initially made use of circular nozzles of 10mm in diameter and then changed to 4mm diameter nozzles (Ndinisa, 2006a). Recently the nozzle size for the new module (SP400) was reduced to below 3 mm while the height of the cassette was raised to 2,600 mm. As a result, SADmhas decreased to 0.3 m3/m2/hr, which falls into the high end for hollow fiber modules (0.1-0.3 m3/m2/hr) found in literature (De Wilde, 2008).
Table 1. Specifi cations of commercially available flat sheet modules (modified from Prieske, 2010).
|Manufacturer/model||Membrane spacing (mm)||Panel height (mm)||SADm(m3/m2/hr)||Superficial gas velocity (m/s)||Aeration|
|A3 Water solutions / M70||7||1050||0.31||0.025||Fine|
The abrupt flow direction change from the down comer to the riser causes significant frictional losses as shown in Fig. 1a (Prieske, 2010). A smoother draft tube edge was introduced to achieve lower bend loss and thus higher circulation velocities (Fig. 1b). An additional acceleration was achieved by locating the aerators at the bottom of the tank instead of at the entrance of the draft tube where they block the available cross section and slow down the flow. Consequently, a much more homogenous bubble distribution across the whole module was achieved, which reduced the chance of the outer channel cloggings. With this configuration, either higher shear forces can be achieved at the same aeration or lower aeration is required to achieve the same liquid velocity. As shown in Fig. 2, 30–50 % faster liquid circulation was observed with the modified frame.
Fig. 1. Enhanced upflow velocity by positioning diffusers outside the frame and adding baffles: a) conventional design, b) improved design (Prieske, 2010).
Fig. 2. Effect of enhanced upflow velocity by positioning diffusers outside the frame and adding baffles – comparison of superficial liquid velocities of conventional and modified designs (Prieske, 2010).
In a recent study, spacers that can increase lateral mixing in between two flat sheet membranes were devised as shown in Fig. 3. This spacer improved membrane performance greatly as shown in Fig. 3, where TMP increase was decreased with the spacers. No details are known yet as of October, 2011.
Fig. 3. Spacer design to improve lateral flow in between two flat sheet membranes (Melin, 2011)
a) Without spacer b) With spacer
Fig. 4. Membrane performance, filtration at 35 LMH, left without, right with spacers. (SAD = specific aeration rate, Nm³/m/h) (Melin, 2011)
© Seong Hoon Yoon